David S. Miller wrote:
I applied patches 1 -> 7, but I had to stop after that.
This patch here will break netlink on my workstation :-)
These "u32" tricks with nlk->groups[0] will not work on
big-endian 64-bit. If unsigned long is 64-bit, you end
up accessing bits 32-63 of nlk->groups[0] in some instances
and then bits 0-31 in other situations.
Please redo this patch, then I'll apply it and the rest.
You only need resend the fixed version of patch 8.
This one should fix the problem. I also redid 9, it was missing a
permission check for joining groups.
[NETLINK]: Support dynamic number of multicast groups per netlink family
Signed-off-by: Patrick McHardy <[EMAIL PROTECTED]>
---
commit 5a78d0310e5039009bceab69f8b4de05b0ae73de
tree dfa621961787223780026b7ffc51eaad85de8424
parent 159b54716bcf5aea4c3141a64072613db7934e11
author Patrick McHardy <[EMAIL PROTECTED]> Mon, 15 Aug 2005 09:22:12 +0200
committer Patrick McHardy <[EMAIL PROTECTED]> Mon, 15 Aug 2005 09:22:12 +0200
net/netlink/af_netlink.c | 69 ++++++++++++++++++++++++++++++++++------------
1 files changed, 51 insertions(+), 18 deletions(-)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -60,21 +60,24 @@
#include <net/scm.h>
#define Nprintk(a...)
+#define NLGRPSZ(x) (ALIGN(x, sizeof(unsigned long) * 8) / 8)
struct netlink_sock {
/* struct sock has to be the first member of netlink_sock */
struct sock sk;
u32 pid;
- unsigned int groups;
u32 dst_pid;
u32 dst_group;
+ u32 flags;
+ u32 subscriptions;
+ u32 ngroups;
+ unsigned long *groups;
unsigned long state;
wait_queue_head_t wait;
struct netlink_callback *cb;
spinlock_t cb_lock;
void (*data_ready)(struct sock *sk, int bytes);
struct module *module;
- u32 flags;
};
#define NETLINK_KERNEL_SOCKET 0x1
@@ -101,6 +104,7 @@ struct netlink_table {
struct nl_pid_hash hash;
struct hlist_head mc_list;
unsigned int nl_nonroot;
+ unsigned int groups;
struct module *module;
int registered;
};
@@ -138,6 +142,7 @@ static void netlink_sock_destruct(struct
BUG_TRAP(!atomic_read(&sk->sk_rmem_alloc));
BUG_TRAP(!atomic_read(&sk->sk_wmem_alloc));
BUG_TRAP(!nlk_sk(sk)->cb);
+ BUG_TRAP(!nlk_sk(sk)->groups);
}
/* This lock without WQ_FLAG_EXCLUSIVE is good on UP and it is _very_ bad on SMP.
@@ -333,7 +338,7 @@ static void netlink_remove(struct sock *
netlink_table_grab();
if (sk_del_node_init(sk))
nl_table[sk->sk_protocol].hash.entries--;
- if (nlk_sk(sk)->groups)
+ if (nlk_sk(sk)->subscriptions)
__sk_del_bind_node(sk);
netlink_table_ungrab();
}
@@ -369,6 +374,8 @@ static int __netlink_create(struct socke
static int netlink_create(struct socket *sock, int protocol)
{
struct module *module = NULL;
+ struct netlink_sock *nlk;
+ unsigned int groups;
int err = 0;
sock->state = SS_UNCONNECTED;
@@ -392,15 +399,23 @@ static int netlink_create(struct socket
module = nl_table[protocol].module;
else
err = -EPROTONOSUPPORT;
+ groups = nl_table[protocol].groups;
netlink_unlock_table();
- if (err)
- goto out;
+ if (err || (err = __netlink_create(sock, protocol) < 0))
+ goto out_module;
+
+ nlk = nlk_sk(sock->sk);
- if ((err = __netlink_create(sock, protocol) < 0))
+ nlk->groups = kmalloc(NLGRPSZ(groups), GFP_KERNEL);
+ if (nlk->groups == NULL) {
+ err = -ENOMEM;
goto out_module;
+ }
+ memset(nlk->groups, 0, NLGRPSZ(groups));
+ nlk->ngroups = groups;
- nlk_sk(sock->sk)->module = module;
+ nlk->module = module;
out:
return err;
@@ -437,7 +452,7 @@ static int netlink_release(struct socket
skb_queue_purge(&sk->sk_write_queue);
- if (nlk->pid && !nlk->groups) {
+ if (nlk->pid && !nlk->subscriptions) {
struct netlink_notify n = {
.protocol = sk->sk_protocol,
.pid = nlk->pid,
@@ -455,6 +470,9 @@ static int netlink_release(struct socket
netlink_table_ungrab();
}
+ kfree(nlk->groups);
+ nlk->groups = NULL;
+
sock_put(sk);
return 0;
}
@@ -503,6 +521,18 @@ static inline int netlink_capable(struct
capable(CAP_NET_ADMIN);
}
+static void
+netlink_update_subscriptions(struct sock *sk, unsigned int subscriptions)
+{
+ struct netlink_sock *nlk = nlk_sk(sk);
+
+ if (nlk->subscriptions && !subscriptions)
+ __sk_del_bind_node(sk);
+ else if (!nlk->subscriptions && subscriptions)
+ sk_add_bind_node(sk, &nl_table[sk->sk_protocol].mc_list);
+ nlk->subscriptions = subscriptions;
+}
+
static int netlink_bind(struct socket *sock, struct sockaddr *addr, int addr_len)
{
struct sock *sk = sock->sk;
@@ -528,15 +558,14 @@ static int netlink_bind(struct socket *s
return err;
}
- if (!nladdr->nl_groups && !nlk->groups)
+ if (!nladdr->nl_groups && !(u32)nlk->groups[0])
return 0;
netlink_table_grab();
- if (nlk->groups && !nladdr->nl_groups)
- __sk_del_bind_node(sk);
- else if (!nlk->groups && nladdr->nl_groups)
- sk_add_bind_node(sk, &nl_table[sk->sk_protocol].mc_list);
- nlk->groups = nladdr->nl_groups;
+ netlink_update_subscriptions(sk, nlk->subscriptions +
+ hweight32(nladdr->nl_groups) -
+ hweight32(nlk->groups[0]));
+ nlk->groups[0] = (nlk->groups[0] & ~0xffffffffUL) | nladdr->nl_groups;
netlink_table_ungrab();
return 0;
@@ -590,7 +619,7 @@ static int netlink_getname(struct socket
nladdr->nl_groups = netlink_group_mask(nlk->dst_group);
} else {
nladdr->nl_pid = nlk->pid;
- nladdr->nl_groups = nlk->groups;
+ nladdr->nl_groups = nlk->groups[0];
}
return 0;
}
@@ -791,7 +820,8 @@ static inline int do_one_broadcast(struc
if (p->exclude_sk == sk)
goto out;
- if (nlk->pid == p->pid || !(nlk->groups & netlink_group_mask(p->group)))
+ if (nlk->pid == p->pid || p->group - 1 >= nlk->ngroups ||
+ !test_bit(p->group - 1, nlk->groups))
goto out;
if (p->failure) {
@@ -887,7 +917,8 @@ static inline int do_one_set_err(struct
if (sk == p->exclude_sk)
goto out;
- if (nlk->pid == p->pid || !(nlk->groups & netlink_group_mask(p->group)))
+ if (nlk->pid == p->pid || p->group - 1 >= nlk->ngroups ||
+ !test_bit(p->group - 1, nlk->groups))
goto out;
sk->sk_err = p->code;
@@ -1112,6 +1143,7 @@ netlink_kernel_create(int unit, void (*i
nlk->flags |= NETLINK_KERNEL_SOCKET;
netlink_table_grab();
+ nl_table[unit].groups = 32;
nl_table[unit].module = module;
nl_table[unit].registered = 1;
netlink_table_ungrab();
@@ -1358,7 +1390,8 @@ static int netlink_seq_show(struct seq_f
s,
s->sk_protocol,
nlk->pid,
- nlk->groups,
+ nlk->flags & NETLINK_KERNEL_SOCKET ?
+ 0 : (unsigned int)nlk->groups[0],
atomic_read(&s->sk_rmem_alloc),
atomic_read(&s->sk_wmem_alloc),
nlk->cb,
[NETLINK]: Add set/getsockopt options to support more than 32 groups
NETLINK_ADD_MEMBERSHIP/NETLINK_DROP_MEMBERSHIP are used to join/leave
groups, NETLINK_PKTINFO is used to enable nl_pktinfo control messages
for received packets to get the extended destination group number.
Signed-off-by: Patrick McHardy <[EMAIL PROTECTED]>
---
commit 773fbe5d7b08299f5a476dbd2ec48e01551348b9
tree 51df6fda3df986f7ba8ea2e432b31bf91a8f7d25
parent 5a78d0310e5039009bceab69f8b4de05b0ae73de
author Patrick McHardy <[EMAIL PROTECTED]> Mon, 15 Aug 2005 09:33:25 +0200
committer Patrick McHardy <[EMAIL PROTECTED]> Mon, 15 Aug 2005 09:33:25 +0200
include/linux/netlink.h | 9 ++++
include/linux/socket.h | 1
net/netlink/af_netlink.c | 95 +++++++++++++++++++++++++++++++++++++++++++++-
3 files changed, 103 insertions(+), 2 deletions(-)
diff --git a/include/linux/netlink.h b/include/linux/netlink.h
--- a/include/linux/netlink.h
+++ b/include/linux/netlink.h
@@ -90,6 +90,15 @@ struct nlmsgerr
struct nlmsghdr msg;
};
+#define NETLINK_ADD_MEMBERSHIP 1
+#define NETLINK_DROP_MEMBERSHIP 2
+#define NETLINK_PKTINFO 3
+
+struct nl_pktinfo
+{
+ __u32 group;
+};
+
#define NET_MAJOR 36 /* Major 36 is reserved for networking */
enum {
diff --git a/include/linux/socket.h b/include/linux/socket.h
--- a/include/linux/socket.h
+++ b/include/linux/socket.h
@@ -272,6 +272,7 @@ struct ucred {
#define SOL_NETBEUI 267
#define SOL_LLC 268
#define SOL_DCCP 269
+#define SOL_NETLINK 270
/* IPX options */
#define IPX_TYPE 1
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -81,6 +81,7 @@ struct netlink_sock {
};
#define NETLINK_KERNEL_SOCKET 0x1
+#define NETLINK_RECV_PKTINFO 0x2
static inline struct netlink_sock *nlk_sk(struct sock *sk)
{
@@ -946,6 +947,94 @@ void netlink_set_err(struct sock *ssk, u
read_unlock(&nl_table_lock);
}
+static int netlink_setsockopt(struct socket *sock, int level, int optname,
+ char __user *optval, int optlen)
+{
+ struct sock *sk = sock->sk;
+ struct netlink_sock *nlk = nlk_sk(sk);
+ int val = 0, err;
+
+ if (level != SOL_NETLINK)
+ return -ENOPROTOOPT;
+
+ if (optlen >= sizeof(int) &&
+ get_user(val, (int __user *)optval))
+ return -EFAULT;
+
+ switch (optname) {
+ case NETLINK_PKTINFO:
+ if (val)
+ nlk->flags |= NETLINK_RECV_PKTINFO;
+ else
+ nlk->flags &= ~NETLINK_RECV_PKTINFO;
+ err = 0;
+ break;
+ case NETLINK_ADD_MEMBERSHIP:
+ case NETLINK_DROP_MEMBERSHIP: {
+ unsigned int subscriptions;
+ int old, new = optname == NETLINK_ADD_MEMBERSHIP ? 1 : 0;
+
+ if (!netlink_capable(sock, NL_NONROOT_RECV))
+ return -EPERM;
+ if (!val || val - 1 >= nlk->ngroups)
+ return -EINVAL;
+ netlink_table_grab();
+ old = test_bit(val - 1, nlk->groups);
+ subscriptions = nlk->subscriptions - old + new;
+ if (new)
+ __set_bit(val - 1, nlk->groups);
+ else
+ __clear_bit(val - 1, nlk->groups);
+ netlink_update_subscriptions(sk, subscriptions);
+ netlink_table_ungrab();
+ err = 0;
+ break;
+ }
+ default:
+ err = -ENOPROTOOPT;
+ }
+ return err;
+}
+
+static int netlink_getsockopt(struct socket *sock, int level, int optname,
+ char __user *optval, int __user *optlen)
+{
+ struct sock *sk = sock->sk;
+ struct netlink_sock *nlk = nlk_sk(sk);
+ int len, val, err;
+
+ if (level != SOL_NETLINK)
+ return -ENOPROTOOPT;
+
+ if (get_user(len, optlen))
+ return -EFAULT;
+ if (len < 0)
+ return -EINVAL;
+
+ switch (optname) {
+ case NETLINK_PKTINFO:
+ if (len < sizeof(int))
+ return -EINVAL;
+ len = sizeof(int);
+ val = nlk->flags & NETLINK_RECV_PKTINFO ? 1 : 0;
+ put_user(len, optlen);
+ put_user(val, optval);
+ err = 0;
+ break;
+ default:
+ err = -ENOPROTOOPT;
+ }
+ return err;
+}
+
+static void netlink_cmsg_recv_pktinfo(struct msghdr *msg, struct sk_buff *skb)
+{
+ struct nl_pktinfo info;
+
+ info.group = NETLINK_CB(skb).dst_group;
+ put_cmsg(msg, SOL_NETLINK, NETLINK_PKTINFO, sizeof(info), &info);
+}
+
static inline void netlink_rcv_wake(struct sock *sk)
{
struct netlink_sock *nlk = nlk_sk(sk);
@@ -1091,6 +1180,8 @@ static int netlink_recvmsg(struct kiocb
netlink_dump(sk);
scm_recv(sock, msg, siocb->scm, flags);
+ if (nlk->flags & NETLINK_RECV_PKTINFO)
+ netlink_cmsg_recv_pktinfo(msg, skb);
out:
netlink_rcv_wake(sk);
@@ -1465,8 +1556,8 @@ static struct proto_ops netlink_ops = {
.ioctl = sock_no_ioctl,
.listen = sock_no_listen,
.shutdown = sock_no_shutdown,
- .setsockopt = sock_no_setsockopt,
- .getsockopt = sock_no_getsockopt,
+ .setsockopt = netlink_setsockopt,
+ .getsockopt = netlink_getsockopt,
.sendmsg = netlink_sendmsg,
.recvmsg = netlink_recvmsg,
.mmap = sock_no_mmap,
